A subject matter expert friend of ours is being pressed ad nauseam to write a book, but his free time is in short supply. It has been suggested to him, again ad nauseam, to  have an LLM do it for him. The book would be generated in days!

Before deferring to a generative AI, some things to keep in mind: “AI Search Has A Citation Problem. We Compared Eight AI Search Engines. They’re All Bad at Citing News,” Columbia Journalism News reported.  “We found that…

• Chatbots were generally bad at declining to answer questions they couldn’t answer accurately, offering incorrect or speculative answers instead.
• Premium chatbots provided more confidently incorrect answers than their free counterparts.
• Multiple chatbots seemed to bypass Robot Exclusion Protocol preferences.
• Generative search tools fabricated links and cited syndicated and copied versions of articles.
• Content licensing deals with news sources provided no guarantee of accurate citation in chatbot responses.

“Because AI chatbots’ responses are dynamic and can vary in response to the same query, the chances are high that if someone ran the exact same prompts again, they would get different outputs.”

What would this do to this subject matter expert’s reputation, should he use copyrighted material without citing the source. It’s called plagiarism, and he is not a plagiarist. Isn’t the creator of the generative AI program itself guilty of plagiarism and where are the repercussions?

“In a post on X, OpenAI CEO Sam Altman said that the company has trained a “new model” that’s “really good” at creative writing,” Tech Crunch reported. “He posted a lengthy sample from the model given the prompt “Please write a metafictional literary short story about AI and grief.” Historically, AI hasn’t proven to be an especially talented essayist.”

As Mark Cuban accurately pointed out at SXSW, “AI is ‘never the answer,’ it’s a ‘tool’ While he admitted that there were some problems with AIs that make mistakes and hallucinate, he noted that human mentors and experts “don’t always get it right, either.”

“Cuban cautioned the crowd not to overly rely on AI. “AI is never the answer. AI is the tool. Whatever skills you have, you can use AI to amplify them,” he said. “AI doesn’t know a good story from a bad story. You need to be creative. AI can do the video — trust me, I can create AI-generated videos. They’re still gonna suck.”

We did warn our potential author friend that all well and good to use a genAI, but to keep in mind that it’s a researcher or intern at best, and not a co-author. In other words, it’s a tool, not a panacea, and that’s a heads up to anyone who believes that any GenAI is capable of delivering a polished final product.

Writing is easy. Editing is where the real work comes in.

And then there’s this: Court filings show Meta staffers discussed using copyrighted content for AI training, said Tech Crunch.

“[M]y opinion would be (in the line of ‘ask forgiveness, not for permission’): we try to acquire the books and escalate it to execs so they make the call,” wrote Xavier Martinet, a Meta research engineer. Martinet floated the idea of buying e-books at retail prices to build a training set rather than cutting licensing deals with individual book publishers. After another staffer pointed out that using unauthorized, copyrighted materials might be grounds for a legal challenge, Martinet doubled down,… “[M]y 2 cents again: trying to have deals with publishers directly takes a long time …”

“Yeah we definitely need to get licenses or approvals on publicly available data still,” (said) Melanie Kambadur (a senior manager for Meta’s Llama model research team), according to the filings. “[D]ifference now is we have more money, more lawyers, more bizdev help, ability to fast track/escalate for speed, and lawyers are being a bit less conservative on approvals.”

Our question: why is it that these employees are collecting salaries? If IP  means nothing, or should be freely available – literally – why is the workforce itself being paid? Turnabout is fair play, after all, no?

Then there’s Amazon, and no doubt et al, who are taking self-published book on its platform, running them through an AI, and turning out cheaper, AI-slop versions of the book. And do keep in mind that no AI-dupe is going to deliver the power in the words and thoughts that an accomplished writer can.

Bottom line/collateral damage: why have an original thought or work of art, written or visual, if it’s going to be appropriated by an AI anyway? Sans adequate or any remuneration for the original artist’s time and talents, of course.

But wait! There’s more! “OpenAI declares AI race “over” if training on copyrighted works isn’t fair use,” Ars Technica reported. “National security hinges on unfettered access to AI training data, OpenAI says,” meaning Sam Altman, and how many times has he just ‘appropriated’ elements he needed, such as using Scarlett Johansson’s voice without properly remunerating or even crediting her. It’s that ‘ask forgiveness’ thing again, and since that didn’t fly in the courts, try scare tactics, yet another fallback tool in Big Tech’s arsenal.

Yet how quick is Big Tech to file lawsuits when it seems another company has appropriated their IP and this just in: “People are using Google’s new AI model to remove watermarks from images,” Tech Crunch reported. “Users on social media have discovered a controversial use case for Google’s new Gemini AI model: removing watermarks from images, including from images published by Getty Images and other well-known stock media outfits.” And note to self: Getty Images are notorious for collecting their vig, so this will be interesting.

It’s a slippery slope and companies who have enormous amounts of dry powder see only their own bottom lines rather than the long-term consequences of freely appropriating original content. AI is a tool and it’s time that founders use it like one, rather than AI dev companies continuing to behave like one.

Onward and forward.